|
Session 22. Multivariate stochastic modelling in finance, insurance and risk management
|
Construction and sampling of nested Archimedean Lévy copulas
|
Oliver Grothe, University of Cologne, Germany
|
The talk is based on the joint work with Marius Hofert
|
|
|
Lévy processes are of growing use in econometric and financial modeling due to their simplicity and their flexibility in modeling tails of distributions. In multivariate models, the dependence structure between the dimensions may be captured by Lévy copulas. However, flexible high-dimensional Lévy copulas are still rare in the literature. We therefore introduce the class of nested Archimedean Lévy copulas. This class models dependencies between Lévy processes hierarchically and overcomes the
inherent symmetry linked to Archimedean Lévy copulas.
Additionally, a fast sampling algorithm for multivariate Lévy
processes with dependence structure specified by Archimedean or nested Archimedean Lévy
copulas is derived from a Marshall-Olkin-type algorithm.
|
|
| Print version |
|
